A central question in the finance field is whether firms have a target capital structure which they attempt to achieve, as well as how capital structure reacts under the influence of
parameters such as profitability, tangibility, the market-to-book ratio, bankruptcy risk, time
period or firm size. Our goal through the whole process of the paper is to understand the
relationship between leverage and profitability in empirical tests of capital structure.
Prominent theories, such as the trade-off, pecking order and market timing theory, are
briefly described over the course of this paper in order to provide a better overview of the
existing ideas and with the goal of comparing empirical results with expected theoretical
predictions. By referring to other papers related to this subject, we present a wider range of
points of view and attempt to verify the statements made in these articles on our own data
sample, consisting of German firms.
Our paper makes an attempt of performing a regression by means of all the parameters
above with a regression model based on the one proposed by Frank and Goyal (2009) and
applied to the given data. This model was chosen not only to ensure comparability with
previous studies on the same subject, but also due to the fact that it accounts for most of the
factors that have been identified in existing literature (e.g. Frank and Goyal, 2003) to play a
significant role in firms. financing decisions.
Supplementary literature which influenced the trend of this paper is offered by articles of
Frank and Goyal (2009), Strebulaev (2003) and Elsas and Florysiak(2008). We follow the
approach proposed in these papers and deduct similar conclusions.
This paper is structured as follows: The influence sources are briefly described under
Literature Review.. The next section, Theory., discusses the main theories present
throughout the paper. Important features of the used dataset are mentioned in the
subchapter Data., with the expected outcomes presented in the subsection Description of
the Regression Analysis.. The results are displayed and interpreted in the Results. section,
while the implications are noted under Conclusion..
II. Literature Review
In this section we summarize the literature that has influenced our theoretical and empirical
work.
The paper by Frank and Goyal (2009) comes as a reaction to the existing literature in the
form of an innovative paper stressing out the fact that commonly used leverage ratios had
been misinterpreted. They examine the relation between profitability on the one hand and
debt and equity issuances on the other, as well as the effect of company size on capital
structure decisions.
Another significant recent contribution to the capital structure research is the work by
Strebulaev (2007), which stresses out the fact that the properties of leverage at refinancing
points differ dramatically in true dynamics and comparative statics.
In their paper, Kayhan and Titman (2007) examine different variables that influence the
capital structure based on historic data. They conclude that although firms behave as though
they have a target debt ratio, they deviate from this target, influenced by different factors,
such as: past profitability, financial deficit, leverage deficit.
Using a dynamic framework and a broad data panel, Drobetz, Pensa and Wanzenried
(2006) suggest that the traditional capital structure theories do not characterize the nature of
the adjustment process towards target debt ratios. They analyze this process by
documenting the effect of firm characteristic variables on the speed of adjustment to target
leverage.
The study by Frank and Goyal (2008) presents the theoretical background of the trade-off
theory and distinguishes between a static trade-off model and a dynamic process of capital
structure adjustment towards a target leverage ratio. In addition to that, it presents several
stylized facts and discusses ways to empirically test these models.
A commonly used method to verify the validity of a static trade-off model would be to
observe the financing behavior of firms, following external shocks to their capital structure,
as analyzed by Myers (1984), who also creates a pecking order model and comes to the
conclusion that the latter better describes the empirically observed relation between
profitability and the leverage ratio.
By comparing predictions of the trade-off theory to the pecking order models, Fama and
French (2002) succeed in finding both analogies and differences between the two; most of
the differences apply to book leverage, but they sometimes affect market leverage as well.
This paper is a confrontation between trade-off and pecking order theories. Using a
completely different approach, Rajan and Zingales (1995) write an empirical paper, which
studies the determinants of firm.s capital structure in the major industrialized countries.
III. Theory
The first step in understanding firm.s capital structure adjustment decisions was made by
the Modigliani-Miller irrelevance proposition in 1958. Before that there was no generally
accepted theory of capital structure. Our work is based on the following three most widely
cited theories: the trade-off, pecking order and market timing theories.
A. Trade-off Theory
The term trade-off is used in different ways by different authors to describe a family of
related theories, some stating that bankruptcy and taxes are being balanced (Kraus and
Litzenberger, 1973), for others it includes agency based arguments (Fama and French,
2002). In all of these theories a manager evaluates the various costs and benefits of
alternative leverage plans, while it is often assumed that an interior solution is obtained so
that marginal costs and marginal benefits are balanced.
The original version of the trade-off theory emerged when corporate income taxes were
added to the Modigliani-Miller irrelevance proposition (1963). This created a benefit for
debt financing by protecting the earnings and profits of firms using the so called tax shield,
with the cost of bankruptcy being the obvious counterbalance. According to Myers (1984) a
firm following the trade-off theory sets a target debt-to-value ratio and then slowly moves
towards this target, which is determined by balancing debt tax shields against costs of
bankruptcy. Further analysis of Myers definition reveals two different forms of the trade-
off theory: the static and dynamic trade-off theories. Accordingly, the static theory can be
defined by a single period model and a trade-off between debt and the costs of bankruptcy,
while the dynamic model states that a firm exhibits an adjustment behavior towards target
leverage and if deviations from this target occur, they are gradually removed over time.
In their paper, Frank and Goyal (2009) highlight the most important implications of the
static trade-off theory, stating that more profitable firms borrow more, repurchase equity
and experience an increase in both book value of equity and market value of equity. Among
the low profit firms there is more variation in the book and market equity. The fact, that
more profitable firms are predicted to have a higher leverage ratio, while empirically they
tend to have lower leverage ratios, is regarded as a defective characteristic of the static
trade-off model.
In order to better understand the effect of profitability on leverage, a dynamic model has
been developed, which allows profitability and leverage to be negatively related in the data
due to various frictions. A number of aspects ignored in a single-period model, are taken
into consideration by the dynamic theory, with the roles of expectations and adjustment
costs being of particular importance. In a dynamic model, the correct financing decision
typically depends on the financing margin that the firm anticipates in the next period.
An important advocate of the dynamic trade-off theory is Strebulaev (2007), who states in
his paper that expected profitability is positively related to leverage at the refinancing
points. However, he further reveals that in a dynamic economy, cross-sectional tests reveal
a negative relation. With infrequent adjustments, an increase in profitability lowers
leverage by increasing future profitability and thus the value of the firm. Similarly, a
decrease in profitability increases leverage.
B. Pecking Order Theory
In the research of firm.s capital structure, the pecking order theory was developed by
Steward C. Myers and Nicolas Majluf in 1984. It states that companies choose their sources
of financing (from internal financing to equity) according to the law of least effort, or of
least resistance. Therefore, internal funds should be the first to be used and when those are
no longer available then debt is issued. The next step, when debt issuance is not enough, is
the equity issuance.
The empirically observed inverse relationship between profitability and debt ratios could be
explained using the pecking order theory. It is an implication of the theory that companies
prefer internal financing; they adapt their dividend payout ratios to their investment
opportunities and try to avoid sudden changes in variable values. Due to the unpredictable
fluctuations in profits and investment opportunities the generated cash flow is not equal to
the capital expenditures. In the case of greater cash flow, the firm pays the debt or invests
in securities; should the cash flow be less than expenditures, then the firm draws down its
cash balance or sells securities, rather than reducing dividends. If external financing is still
required, then the firm will start by choosing debt, followed by hybrid securities
(convertible bonds) and then equity as a last resort. As a supplementary argument, issuing
costs are lowest for internal funds, low for debt and highest for equity. There is also the
negative signaling to the stock market associated with issuing equity and positive signaling
associated with debt.
Several studies on this topic have demonstrated that the pecking order theory is a good
approximation of reality. If we were to consider the conclusions of Fama and French (2002)
and several other authors, some features of their data were better explained by this theory
than by the classic trade-off theory. Frank and Goyal (2007) on the other hand critically
underline the fact that the pecking order theory fails where it should actually hold, namely
for small firms where information asymmetry is presumably an important issue. Other
critical voices like An alternate test of Myer.s pecking order theory of capital structure:
the case of South Korean firms by Ang and Jung (1993), run local studies which fail to
support the Myers.s pecking theory when interpreting the results on marginal and
sensitivity analyses.
C. Market Timing Theory
The market timing theory states that firms prefer external equity when the cost of equity is
low and debt when the cost of equity is high. Similar to the pecking order theory, the
market timing theory does not recognize the existence of a target capital structure as stated
by Huang and Ritter (2004). To make this suggestion even more solid, Baker and Wurgler
(2002) note that even if there was an optimum, the benefits of market timing are perceived
to be higher than the costs of deviating from this optimum.
While in both the pecking order and market timing theories managers are primarily
concerned about current shareholders, the latter rejects the necessity of the existence of a
rational expectations equilibrium, stating instead that managers may believe that the equity
of their firm is over- or undervalued by the market (Baker/Wurgler, 2002). As pointed out
in the same paper, this suggestion does not require an inefficient market or correct
predictions on behalf of the managers, only the fact that they believe they can time the
market.
The primary prediction made by the market timing theory is that as long as managers
believe their stock is overvalued, they tend to issue more equity. Conversely, Baker and
Wurgler (2002) observe that equity is repurchased or debt issued when the stock is
undervalued. Another important point in this context as seen in Huang and Ritter (2004) is
that firms may resort to equity or debt issuances even if they have no immediate financing
needs. This is explained by the fact that issuing overvalued securities is in itself a project
with positive net present value. Baker and Wurgler (2002) further indicate that temporary
fluctuations in the value of a firm can lead to permanent changes in its capital structure,
which is determined to a large extent by past financing decisions. In addition, there is no
target ratio which firms would aim to achieve.
The predictions of the market timing theory are supported by empirical evidence. Firstly, as
discovered by a survey of Graham and Harvey (2001), sixty seven percent of CFO.s
interviewed agree that over- or undervaluation of stock is an important factor when
considering issuing equity. Secondly, as found by Baker and Wurgler (2002) as well as
Loughran, Ritter and Rydqvist (1994) among others, there is a strong positive relationship
between the market-to-book ratio, which is used as an indicator for perceived over- or
undervaluation of a firm.s stock, and the propensity to issue equity. Nevertheless, the
market timing theory finds no precise connection between the leverage ratio and
profitability, which is why we do not consider the implications of this theory in our
empirical test.
IV. Model
A. Data
The sample we used in our paper comes from the annual financial statement data available
in Hoppenstedt for German exchange listed firms with unregulated capital structures,
excluding financials and utilities. The observation period for the sample is from 1987 to
2006 and the accounting standards used are: IAS, US-GAAP, "HGB-
Gesamtkostenverfahren", and "HGB-Umsatzkostenverfahren". For firms with both a
consolidated and an individual financial statement, the consolidated financial statement has
been selected. Market data comes from the Datastream database by Thomson Financial.
Financial statement data are deflated using the GDP-deflator with base year 2005. The
following variables have been used throughout the paper and in the capital structure
regression. We use the same data as Elsas and Florysiak (2008).
Distribution
Variable
N
Mean
SD
Min
Max
25th
50th
75th
99th
Debt ($ millions)
10082
1103.623
7226.356
0.06
145963.00
12.25
52.0
208.25
145963
Book equity ($ millions)
10082
394.616
2054.007
0.00
36801.00
9.75
31.5
114.50
36801
Market equity ($ millions)
10082
821.511
3963.241
0.59
67071.00
18.00
61.0
246.00
67071
Assets ($ millions)
10082
425.246
2713.176
0.00
67534.00
2.25
19.5
89.75
67534
Size
10082
18.2712
2.413945
10.31
25.70
16.81
18.4
19.73
24
Book leverage
10082
0.596
0.225
0.02
1.00
0.454
0.631
0.763
1
Market leverage
10082
0.452
0.254
0.00
0.97
0.246
0.450
0.654
0.945
Median Ind. Leverage
10082
0.430
0.140
0.03
0.83
0.348
0.451
0.529
0.692
Profitability
10082
0.093
0.159
-1.74
0.68
0.044
0.105
0.166
0.467
Tangibility
10082
0.286
0.234
0.00
1.20
0.103
0.243
0.407
0.955
Market-to-book
10082
2.179
4.290
0.39
97.26
1.053
1.331
1.923
18.222
Variable definitions:
Debt = Long-term debt + Short-term debt
Book equity = Common shareholder equity
Market equity = Outstanding shares X Closing share price
Assets = Book assets
Book leverage = Debt / (Debt + Book equity)
Market leverage = Debt / (Debt + Market equity)
Profitability = EBITDA/Total assets
Tangibility = Net property plant and equipment / Total assets
Market-to-book = Market value of assets / Book assets
Median Industry Leverage = Median market debt ratio at the time t, following Fama and
French (1997) industry classification
B. Description of the Regression Analysis
In order to assess the importance of several factors in determining the capital structure of
firms, a regression analysis can be utilized with leverage as the dependent variable.
The first step to this end is to identify which factors are particularly useful in explaining
how firms make their financing decisions and in what ways firms differ in their capital
structures. A paper published by Frank and Goyal in 2003 examines the influence of
several variables in the financing decisions of US publicly traded firms. A primary finding
of this paper is that the level of leverage in the industry in which the firm operates, the
market-to-book ratio of the firm, the size of the company as measured by the natural
logarithm of its sales, bankruptcy risk as measured by Altman.s Z-Score and the tangibility
of its assets are all significantly correlated with the firm.s leverage ratio. All this is
consistent with existing literature and widely used in similar regression analyses. Another
interesting finding of the same study is that the effect of corporate profits on leverage is not
particularly robust (though still negative, contrary to the predictions of the static trade-off
theory).
Now that several influencing factors have been identified, the next step can be taken, which
is to specify a regression. In order to ensure comparability with previous studies on the
same subject we try to set up a regression with similar characteristics to the ones widely
used in existing literature. Therefore we utilize a regression similar to the one specified in a
paper of Frank and Goyal (2009). The merit of this estimation is the use of leverage ratios,
something that is common in the previous literature. The regression model is as follows:
Included in the above model are all factors previously mentioned except of the bankruptcy
risk of a firm due to the lack of sufficient data in the database we used. In addition, a year
dummy has been included.
Using the above regression, we first try to formulate predictions as to what should be
expected according to capital structure theory. The median leverage of an industry should
be positively related to the leverage ratio of a firm in a trade-off context. This is justified by
the fact that firms in the same industry are, to a large extent, exposed to similar threats and
have to adapt to similar developments (Frank/Goyal, 2003). As further stated in the same
study, under the pecking order theory only an indirect link between the two is recognized,
which exists only to the extent that the industry median leverage serves as a proxy for the
firm.s financing needs.
The market-to-book ratio is expected to be negatively correlated with leverage, both in a
trade-off and in a pecking order context. According to the trade-off theory, this stems both
from the need to retain growth options and the fact that growth firms are riskier, losing
more value when they go into distress. The pecking order theory on the other hand suggests
that more profitable firms make more use of internal cash flows than of external forms of
financing, such as debt (Frank/Goyal, 2003).
According to trade-off theory, tangible assets are expected to have a positive effect on
leverage. This is attributed to the fact that assets like property and equipment are usually
easier to value compared to intangible assets, meaning that they can easily be used as
collateral (Frank/Goyal, 2003). In the context of the pecking order theory on the other hand,
tangibility is expected to have a negative correlation with leverage. Tangibility is associated
with lower information asymmetry, something that makes equity comparatively less costly
(Frank/Goyal, 2007).
Theory suggests that a company.s size is another important determinant of capital structure
decisions. We use the natural logarithm of sales as a measure for firm size, which seems to
be a better indicator than log of assets, as has been empirically shown by Frank and Goyal
(2003). The static trade-off theory suggests that larger companies are usually able to pile up
more debt, due to the fact that they are generally more diversified and have a lower default
risk. In addition, they usually have existed for a longer period of time and are well known
to debt markets, meaning that their agency costs of debt are lower (Frank/Goyal, 2003). All
these factors lead to the unambiguous conclusion, that in a trade-off context, firm size is
positively related to leverage (Frank/Goyal 2003 and 2007). The pecking-order theory on
the other hand, does not offer any clear predictions. It is often interpreted as predicting an
inverse relation between the two, something that is attributed to the lower level of volatility
in large firms. assets. This lower level of volatility makes equity issuances relatively
cheaper compared to smaller, more volatile firms (Frank/Goyal, 2007). However, this effect
is countered by the costs of adverse selection of existing assets, an effect which, due to the
large volume of the assets, may be significant. If the latter effect outweighs the former, then
a positive relation between leverage and profitability can be predicted, as noted by Frank
and Goyal in 2003 and 2007.
The effects of profitability on corporate leverage are also complex. The static trade-off
theory points to a positive relation between profits and leverage, as more profitable firms
can issue more debt and have more profits to shield from taxation (Frank/Goyal, 2009).
This positive relation between the two should be more emphatic in the case of book
leverage (Fama/French, 2002). On the other hand, the relation between market leverage and
profitability cannot be so easily predicted. This stems from the fact that profits can also be
used as a proxy for firm growth, as stated by Frank and Goyal in 2007, with higher profits
leading to an increase in the market value of the firm. If this effect is strong enough, then an
inverse relation between profitability and market leverage could be observed. The pecking
order theory, in the contrary, directly implies that increased profitability should result in
less external financing and hence, lower leverage, as suggested by Frank and Goyal (2007).
The market timing theory makes no clear predictions, as financing decisions also rely on
the current state of the market and the expectations of managers (Frank/Goyal, 2003 and
Baker/Wurgler, 2000).
5. Results
In this part, we present the results of the regression analyses we performed on the examined
panel data, as described in part IV. We first present the results of the leverage regressions
performed on the whole sample.
Table I
Time-Series Regressions for Book and Market Leverage
The sample consists of 1182 German firms with observations ranging in the period from
1987 to 2006. The table presents estimates of the following leverage ratio regression,
which has been derived from the seminal 2009 paper of Frank and Goyal:
whereLeveraget is the ratio of debt over debt plus book equity in column (1) and the ratio of
debt over debt plus market equity in column (2). The explanatory variables, which include
profitability, median leverage of the industry (IndMedianLev), the market-to-book ratio
(
, tangibility and size, are described in part IV and lagged by one year. The
regressions include year fixed effects. The coefficients have been rounded to 4 decimal
digits. The t-statistics are reported in parentheses below the respective coefficients. *:
Significant at the 10 percent level. **: Significant at the 5 percent level. ***: Significant at
the 1 percent level.
Explanatory variables
Book Leverage
(1)
Market Leverage
(2)
Profitabilityt-1
-0.1964***
(-15.48)
-0.2776***
(-19.61)
IndMedianLevt-1
0.0558***
(4.10)
0.0804***
(5.29)
M/Bt-1
-0.0032***
(-8.13)
-0.0066***
(-15.04)
Tangibilityt-1
0.1378***
(10.77)
0.0794***
(5.56)
Sizet-1
0.0263***
0.0536***
(12.72)
(23.23)
Constant
0.0849**
(2.21)
-0.5296***
(-12.37)
R2 Overall
0.1729
0.2247
Observations
8802
8802
The estimates of the above regressions generally comply with those observed in similar
analyses. In particular the results are overly similar to those reported in the paper of Frank
and Goyal (2009), from which the regression has been derived.
The coefficients for profitability are robustly negative for both book and market leverage,
something that directly contradicts the static trade-off theory, but seems to agree with the
pecking order theory. The findings are in line with those reported by Frank and Goyal
(2009) and Fama and French (2000). However, Frank and Goyal (2009) further elaborate
on this issue, offering an alternate explanation about the relation between leverage and
profitability.
Industry median leverage is found to have a highly significant positive effect on both book
and market leverage, a fact that supports the predictions of the trade-off theory and does not
contradict the pecking order theory. Frank and Goyal (2003) state that the median industry
leverage is one of the strongest and most consistent predictors of leverage.. In their paper
of 2009, they report a strongly positive effect on leverage. Our estimation has shown a
positive effect, significant at the 1 percent level, but the coefficients are lower than those
reported in the aforementioned study of Frank and Goyal. However, this difference could
be attributed to other factors such as the smaller sample size employed in our study or the
nature of German firms, in contrast to the behavior of their American counterparts.
The market-to-book ratio negatively affects leverage. The result is statistically significant,
consistent with the predictions of both the trade-off and the pecking order theories as well
as in line with the findings of Frank and Goyal (2009)
The degree of asset tangibility positively influences leverage. This finding is consistent
with the predictions of the trade-off theory, but seems to contradict the pecking order
theory. Frank and Goyal (2009) report similar results.
The estimated coefficients for firm size as measured by log of sales are positive and highly
significant for both book and market leverage. This is what was expected in a trade-off
context. Under a pecking order perspective the effect is not clearly predicted. Nevertheless,
the results make the case for the higher significance of the costs of adverse selection on
existing assets compared to the relative attractiveness of equity, due to lower volatility. Last
but not least, our findings are consistent with those reported by Frank and Goyal (2009) and
Fama and French (2000).
Table II
Time-Series Regressions for Book and Market Leverage sorted by profitability
Table II presents estimates of the previously discussed leverage ratio regression sorted by
profitability in ascending order and divided into the 25 and 75 percentiles, where
Leveraget is the ratio of debt over debt plus book equity in column (1) and the ratio of debt
over debt plus market equity in column (2). The explanatory variables, which include
profitability, median leverage of the industry (IndMedianLev), the market-to-book ratio
(
, tangibility and size, are described in part IV and lagged by one year. The
regressions include year fixed effects. The coefficients have been rounded to 4 decimal
digits. The t-statistics are reported in parentheses below the respective coefficients. *:
Significant at the 10 percent level. **: Significant at the 5 percent level. ***: Significant at
the 1 percent level.
Explanatory
variables
Book Leverage
Market Leverage
25th
75th
25th
75th
Profitabilityt-1
-0.1389***
-0.1658***
-0.1403***
-0.2443***
(-5.38)
(-4.51)
(-6.04)
(-6.45)
IndMedianLevt-1
0.2601***
0.0072
0.5168***
0.2476***
(7.74)
(0.28)
(17.09)
(9.30)
M/Bt-1
0.0008
0.0015
-0.0049***
-0.0140***
(1.32)
(1.34)
(-8.53)
(-12.15)
Tangibilityt-1
0.3230***
0.1983***
0.2352***
0.1187***
(8.64)
(8.84)
(6.99)
(5.14)
Sizet-1
-0.0015
0.0222***
0.0271***
0.0299***
(-0.34)
(5.00)
(6.82)
(6.53)
Constant
0.4059***
0.1169
-0.2644***
-0.2504**
(5.36)
(1.37)
(-3.88)
(-2.85)
R2 Overall
0.1171
0.0932
0.4092
0.2575
Observations
2521
2521
2521
2521
When sorting the sample by profitability and applying the leverage ratio regression on the
25th and 75th percentiles, we observe some patterns for the most and least profitable firms,
but these findings match our previous results and the results obtained by Frank and Goyal
(2009), with some minor variances.
The outputs for the profitability coefficient are negative and highly significant, supporting
our findings from the previous regression, but also contradicting the static trade-off theory.
By splitting the sample into two categories and analyzing the regression results, we can
observe that the profitability has a stronger negative effect over the leverage ratio of more
profitable firms while the effect over the least profitable firms is almost equal in both cases.
While the differences are not very significant, it shows that more profitable firms use their
additional income to pay off debts and thus reduce their leverage ratio. This is an effect
explained by the pecking order theory.
According to Frank and Goyal (2003), the median industry leverage is one of the most
reliable factors in the leverage decisions of publicly traded U.S. firms, with positive effects
on the leverage ratios. Here we observe highly significant results for the 25th percentile for
both leverage ratios and higher positive values than in our previous findings. The median
industry leverage has a strong effect over the leverage ratios of less profitable firms, due to
the fact that these firms are more sensitive and exposed to changes in industry leverage,
with the market leverage ratio experiencing the strongest effects. The data for the more
profitable firms display a low statistical significance, with lower values than that of 25th
percentile, because the leverage ratio of these firms is not so exposed to variances in the
industry leverage.
The market to book ratio is also considered reliable by Frank and Goyal (2003), with a
negative effect on the leverage ratio, which is also consistent with the findings by Rajan
and Zingales (1995), who show that the negative relation between leverage and market-to-
book exists in all G7 countries. Even if the output of the book leverage regression is
positive, it displays small and statistically insignificant numbers. With increasing market
value of the assets, the market leverage decreases, an effect which is stronger for the more
profitable firms.
The tangibility variable does not show any significant deviation from the previous findings,
having a positive influence over the leverage ratios. Firms in the 1st quartile experience a
higher influence of the tangibility ratio over their leverage, because an increase in
tangibility means greater expenditures and the less profitable firms have to use debt more
often than the more profitable companies.
The regression output for the size variable delivers values that are in line with our previous
findings and the findings provided by other studies, excluding the 1st quartile of the book
leverage, which shows a negative but statistically insignificant result. The values for the
other categories do not show significant differences, due to the fact that the size of the
company has roughly the same effect over their leverage ratio, independently of their
profitability.
By conducting two different regressions, one using the original sample and the other by
grouping the data, we can properly understand how each variable is influencing the capital
structure ratios. Of particular interest was the relationship between the leverage ratios and
the profitability, where the results match the findings obtained by Frank and Goyal (2009)
and also correspond with the pecking order theory. It is also noticeable that the variables
that are negatively influencing the leverage ratios, respectively profitability and market to
book ratio, are displaying stronger effects over the more profitable firms, while the less
profitable firms are more sensitive to changes of the positive variables, namely the median
industry leverage and tangibility, with the size coefficient being roughly equal for both
groups.
6. Conclusion
The goal of this paper was to provide an overview and analyze the ambiguous relationship
between leverage and profitability in empirical tests of capital structure. One selected
empirical test was analyzed and replicated, namely the regression analyses used by Frank
and Goyal (2009), to preserve the link to the existing literature on this subject.
We additionally split the sample into the 1st and 3rd profitability quartiles and perform the
two leverage ratio regressions, noticing some interesting patterns and observing more
details of the impact that various variables have on the leverage ratio.
We come to the conclusion that profitability has a negative effect on the book and market
leverage ratios, while the more profitable firms experience a stronger impact, due to the fact
that this increase in profitability is usually leading to a decrease in debt and an increase in
market value of assets.
This negative relationship is a contradiction to the static trade-off theory, which states that
more profitable firms tend to issue debt and repurchase equity, which would lead to a
positive relationship. A further study of the trade-off theory was done by Strebluaev (2007),
who elaborates on the dynamic trade-off theory by considering various frictions which are
ignored in the classic model.
Our observations match those of the specialized literature, which support the predictions of
the pecking order theory. Frank and Goyal (2009) offer additional explanations in this
regard, stating that the commonly used leverage ratios may be misleading due to effect of
already acquired debt and therefore they are usually misinterpreted. From our results, we
conclude that firms have no target leverage ratio as predicted by the trade-off theory.
However we belive that further research has to be conducted whereas the paper by Frank
and Goyal (2009) can be used as a starting point.
